Skoči na glavni sadržaj

Izvorni znanstveni članak

https://doi.org/10.31534/engmod.2026.1.ri.02f

Optimized PV-Wind-Storage System for Reduced Losses

Amina Gabour orcid id orcid.org/0000-0003-3145-7153 ; LGMM Laboratory, Department of mechanical engineering, Faculty of technology, University 20th August 1955 Skikda, ALGERIA *
Abderraouf Bouakkaz ; LGMM Laboratory, Department of mechanical engineering, Faculty of technology, University 20th August 1955 Skikda, ALGERIA
Toufik Sebbagh orcid id orcid.org/0000-0002-1656-0681 ; LGMM Laboratory, Department of mechanical engineering, Faculty of technology, University 20th August 1955 Skikda, ALGERIA
Kamelia Boucenna orcid id orcid.org/0000-0002-2427-6031 ; LGMM Laboratory, Department of mechanical engineering, Faculty of technology, University 20th August 1955 Skikda, ALGERIA

* Dopisni autor.


Puni tekst: engleski pdf 513 Kb

verzije

str. 19-33

preuzimanja: 3

citiraj

Preuzmi JATS datoteku


Sažetak

The main focus of this study is on optimizing an Autonomous Energy System (AES) composed of Photovoltaic (PV) generators, a Wind Turbine (WT) generator, and an Energy Storage System (ESS). Moreover, the main objective aims to define the optimal system requirements for each subsystem to minimize total wasted energy, thereby improving the global efficiency and autonomy of the system. Frequently, a significant amount of renewable energy is wasted when demand is fully satisfied, and the storage system is completely charged, thus leading to underutilization of available resources. Most importantly, the optimal configurations of PV, WT, and ESS were determined in this study to minimize total energy losses. Besides, the simulation considered local conditions, including solar irradiation, ambient temperature, wind speed, and the energy demand profile of a city in Skikda, Algeria. In addition, we carried out an optimization using a Particle Swarm Optimization (PSO) algorithm implemented in MATLAB. The results show that the application of the loss-minimization approach effectively identifies the optimal capacity for each system component, thereby reducing total energy losses and improving the global efficiency and utilization of the energy system accordingly.

Ključne riječi

optimization, energy losses, renewable energy, autonomous energy system, PSO

Hrčak ID:

349571

URI

https://hrcak.srce.hr/349571

Datum izdavanja:

17.7.2026.

Posjeta: 12 *




1. INTRODUCTION

Categorically, autonomous energy systems, integrating renewable energy sources with Energy Storage Systems (ESS), have emerged as a practical and sustainable solution for supplying electricity to isolated and off-grid sites. Therefore, these systems are particularly relevant in remote areas where grid extension is technically challenging or economically prohibitive. Besides, by exploiting locally available renewable resources, such as solar and wind energy, hybrid systems substantially reduce reliance on fossil fuels, mitigate greenhouse gas emissions, and strengthen energy security, while ensuring a continuous and reliable power supply[1].

Certainly, photovoltaic (PV)–wind hybrid systems combined with ESS are particularly attractive among various configurations proposed in the literature due to the complementarity of solar and wind resources. Generally, solar energy availability has been shown to be higher during the day and during summer periods, whilst wind energy can be more abundant at night or during different seasons. The integration of ESS further improves system performance by mitigating the intermittency of renewable generation, balancing supply and demand, and enhancing system reliability[2]. Nonetheless, the benefits of such hybrid systems strongly depend on their design and operational strategy.

Actually, the implementation of a hybrid PV–wind–ESS system requires careful consideration of several design parameters, including the number, size, and type of each component, as well as their operational constraints. Additionally, an inappropriate configuration may lead to excessive energy losses, higher investment and operating costs, reduced component lifetime, and poor overall system efficiency. Therefore, optimal system sizing and configuration are crucial for maximizing energy production, minimizing costs, reducing energy losses, and ensuring long-term system sustainability.

More importantly, the configuration and optimization of hybrid energy system have attracted significant research interest over the past decades[3,4]. In fact, various optimization methods have been proposed with the main goal of minimizing the total system costs while satisfying reliability and technical constraints. As for the classical approaches, such as linear programming[5], response surface methodology[6,7], dynamic programming[8] and the simplex algorithm[9], they have been widely applied. In parallel, metaheuristic and intelligent optimization techniques have gained increasing popularity due to their ability to handle nonlinear, multi-objective, and complex optimization problems. Subsequently, these include teaching–learning-based optimization (TLBO)[10], genetic algorithms (GA)[11], particle swarm optimization (PSO)[12], tabu search (TS)[13], simulated annealing (SA)[14], and hybrid and combined between optimization methods[15,16]. In order to assess the performance of hybrid energy systems and guide the optimization process, several criteria and indices have been adopted in the literature[17]. Moreover, commonly used performance indicators include: loss of power supply probability (LPSP), levelized cost of energy (LCOE), cost of energy (COE), renewable energy fraction (REF), system reliability indices, and energy losses. Despite the fact that economic and reliability criteria have been extensively studied, energy losses—arising from power conversion, storage inefficiencies, and mismatches between generation and load—remain a key factor that directly affects system efficiency, operational performance, and component degradation. In fact, this paper proposes the direct minimization of total energy losses as the sole optimization objective for a PV–Wind–ESS hybrid autonomous system, unlike dominant approaches based on economic criteria (LCOE, COE) or reliability indices (LPSP). Besides, the objective function integrates three components: (i) wasted surplus energy (renewable generation exceeding both load demand and storage capacity), (ii) battery losses from charge/discharge cycles and self-discharge, and (iii) conversion losses in the DC/DC converters. This approach targets the physical inefficiencies of the system regardless of cost or demand satisfaction which some cost-based or reliability-based objectives alone cannot capture. This improves overall efficiency, reduces component stress (extending service life), and maximizes the use of available renewable resources. Also, this loss-minimization framework, applied to the optimal sizing of a PV–Wind–ESS system in the Algerian context, represents, to the authors' knowledge, an original contribution to the literature.

The remaining part of this paper is organized as follows: Section 2 provides a detailed modelling of the system under study, including the main components: PV, wind turbine and an energy storage system, along with the presentation of the power balance equations; Section 3 introduces the mathematical formulation adopted for the optimization problem; Section 4 introduces the case study of the autonomous energy system located in Skikda (Algeria), providing specification of the meteorological data (solar radiation, temperature and wind speed), load profile, and characteristics of the components implemented in the simulations; Section 5 provides presentation and discussion of the main results. As for the conclusion, it was accordingly included in Section 6.

2. SYSTEM MODELLING

2.1 System Configuration

Figure. 1 demonstrates the proposed system in this study. The power system consists of photovoltaic generators, Wind turbine and a battery bank, accordingly.

image1.png

Fig. 1 Configuration of a PV–WT–ESS system

2.1 PV System

The output power of a PV panel (PPV) can be written as follows[18]:

PPV(t)=(PSTCGGSTC(1−γ(Tc−Tc,STC))) (1)

Tc(t)=Ta+GGNOCT(NOCT−Ta,NOCT) (2)

where:

Tc is the obtained cell temperature;

PSTC, GSTC and Tc,STC, are the rated power, solar radiation and the cell temperature under standard test condition (STC) GNOCT and Ta,NOCT is solar radiation and the ambient temperature under nominal operating cell temperature (NOCT), whilst NOCT is a constant.

Each time, the net output power of PV generator t is equal to:

PPVnet(t)=PPV(t)ηCONV (3)

ηCONV represents the efficiency of the DC/DC power converter, the value of which is 0.96.

2.2 Wind System

The output power of the wind turbine (Pw) can be calculated through the following formula[17]:

Pw(t)={Pn.(V2−Vd2Vn2−Vd2)ifVd≤V(t)≤VnPnifVn≤V(t)≤Vc0ifV(t)≤VdouV(t)≥Vc (4)

where:

Pn represents the rated power of the turbine;

V, Vn,Vdand Vcare respectively wind speed, nominal wind speed, cut-in wind speed and cut-out wind speed.

The net output power of the wind turbine Pwnet(t) when t is equal to:

Pwnet(t)=Pw(t)ηCONV (5)

ηCONV represents the efficiency of the DC/DC power converter, whose value is 0.96.

2.3 Battery Bank

The power of battery,PBnet(t), accounting for energy losses, which can be calculated using the following formula[18]:

PBnet(t)=PBloss(t)+PBself−disc(t)+PB(t) (6)

WB(t+1)=WB(t)+PBnet(t)Tsim (7)

where:

PBnet(t) is the net power from / to the battery at t, and PB(t) is the power from / to the battery at t, Tsim is the simulation time of 1 hour.

The Equation (8) shows the self-discharge losses of the battery:

PBself−disc(t)=WB(t−1)σ100Tsim (8)

where WB(t−1) is the battery energy at a previous instant, and σ is the coefficient of self-discharge expressed in percent with respect to the battery energy per hour assumed to be 0.004%.

Concerning the battery efficiency-related losses, they were modelled using an efficiency coefficient ηB which includes the converter, charging and discharging efficiency (it was assumed that ηB=0.9). Moreover, the equations used to model these losses depend on the direction of the power flow, to be precise, charging Eq. (9) or discharging Eq. (10) process.

PBloss(t)=PBnet(t)(1−ηB)=(PB(t)+PBself−disc(t))1−ηBηB (9)

PBloss(t)=−PBnet(t)1−ηBηB=−(PB(t)+PBself−disc(t))(1−ηB) (10)

In this model, a power limit of maximum charge and discharge power is considered as a constraint; to be exact, the power limit of charging Eq. (11) and the power limit of discharging Eq. (12) can be calculated as follows:

PBmax.in(t)=(QBVB−WB(t−1))ηBTsim (11)

PBmax.out(t)=1Tsim(WB(t−1)−QBVBSOCMINηB) (12)

where PBmax.in, PBmax.out present the maximum power of charging and discharging of the battery at t, as QB is the total capacity of the battery and VB is the battery voltage.

2.4 Power Balance Model

The balance of power in each period (t) is presented by the following equation[18]:

PPVnet(t)+Pwnet(t)+PB(t)−Pwasted(t)−(PLOAD(t))ηINV=0 (13)

where PPVnet(t) presents the net power output of the PV production,PB is the charged and discharged power of the battery, Pwasted refers to the excess of power due to the power of PV and the wind turbine output when the battery is fully charged and at the time of satisfaction of the demand, PLOAD corresponds to the power demand, and ηINV refers to the efficiency of DC-AC converter used in this system.

Regarding the power balance model, the following constraints should be respected.

PB(t)<PB,MAX (14)

SOCMIN≤SOC(t)≤SOCMAX (15)

Where PB,MAX is the maximum charging and discharging power of the battery, SOC,SOCMIN and SOCMAX are the state of charge at t, the minimum state of charge and the maximum state of charge, respectively.

3. PROBLEM DESCRIPTION AND MATHEMATICAL FORMULATION

3.1 Power Balance Model

In this paper, the objective function is to minimize the total energy losses in the system, including wasted energy when the demand is satisfied and the battery is fully charged, as the aim is to optimize the utilization of energy of PV, wind turbine and battery.

f=min∑t=18760PTotalloss(t) (16)

PTotalloss(t)=Pwasted(t)+PBloss(t)+PCONloss(t) (17)

(PCONloss(t)=PPV(t)+Pw(t))−(PPVnet(t)+Pwnet(t)) (18)

Where PCONloss presents the power conversion loss in the system.

3.2 PSO Optimization Algorithm

Particle Swarm Optimization is a metaheuristic optimization technique inspired by the collective social behaviour of flocks of birds and schools of fish, originally introduced by Kennedy and Eberhart[19]. The algorithm simulates the movement of a population of particles within a multidimensional search space. Each particle represents a potential solution and is defined by its position X(k) and velocity V(k) at iteration 𝑘. The PSO process starts with randomly initialized particle positions and velocities. During each iteration, particles adjust their trajectories taking into account both their individual best-known position and the global best position found by the swarm. The velocity and position of each particle are updated according to the following equations:

Vi(k+1)=w.Vi(k)+c1r1(Xi,best(k)−(Xi(k))+c2r2(Xg,best(k)−(Xi(k)) (19)

Xi(k+1)=Xi(k)+Vi(k+1) (20)

Where, Xi,best denotes the personal best position of particle i, and Xg,best represents the global best position of the swarm, corresponding to the best solution obtained so far. Further, the coefficients c1 and c2 are the cognitive and social acceleration factors, respectively, whilst r1and r2 are uniformly distributed random numbers in the range [0, 1]. The inertia weight w controls the balance between exploration and exploitation and is updated dynamically as follows:

w(k)=wmax−(wmax−wmin)Niter (21)

Where wmax and wmin are the maximum and the minimum inertia weights, respectively, and Niter is the total number of iterations.

The selected PSO parameters adopted in the proposed scheduling strategy are summarized in Table 1.

Table 1 The input parameters of the PSO algorithm

Parameter Value
wmin 0.4
wmax 0.9
C1and C2 2.05
Variables [PPV, PWT, CBAT]
Lower bound [0, 0, 2000]
Upper bound [5000, 5000, 400000]
Population 30
Max iteration 500

4. CASE STUDY

Most importantly, this section presents the application of the optimization model to the proposed autonomous energy system (AES), composed of photovoltaic generators (PV), a wind turbine generator (WT), and a battery energy storage system (ESS). The objective is to minimize total energy losses and determine the optimal system capacity to satisfy the annual energy demand.

4.1. Components’ Characteristics

The technical characteristics of the system components are presented in Table 2 below.

Table 2 Components’ Characteristics

Parameter Value
Solar radiation at STC 1000 W/m²
Cell temperature at STC 25 °C
Solar radiation at NOCT 800W/m²
Ambient temperature at NOCT 20°C
Rated wind velocity Vr 12.5m/s
Cut in wind velocity 2 m/s
Cut out wind velocity 42 m/s
The inverter and converter efficiency 96 %
Battery efficiency ηB 90%
Battery minimum state of charge SOCMIN 20%
Battery maximum state of charge SOCMAX 100%
Battery self-discharge losses 0.04%/day

4.2. Meteorological Data and Load Profile

The hybrid energy system is assumed to be installed in Skikda, Algeria (latitude 36°56′ N, longitude 6°57′ E, altitude 24 m above sea level), and is designed to supply the electricity demand of remote consumers.

Accordingly, the solar radiation, temperature and wind speed are illustrated in Figure 2, whilst the load profile is shown in Figure 3.

image2.png

Fig. 2 Meteorological Data of PV–Wind used in this study

image3.png

Fig. 3 Load demand profile used in this study

5. RESULTS AND DISCUSSION

To evaluate purpose of the robustness of the optimization process of PSO, we performed five independent optimization runs using different random initializations, as presented in Table 3. The obtained results were highly consistent, presenting nearly identical optimal solutions in all runs.

The computational time was also recorded for each run. In addition, the execution time ranged from 933 s to 964 s, with an average value of approximately 946 s per optimization run. All simulations were performed on a computer equipped with an 11th Gen Intel® Core™ i7-11850H processor operating at 2.50 GHz and 16 GB of RAM.

The obtained results are summarized in Table 3 and Table 4 below.

As shown in Table 3, the PSO optimization determined the optimal component sizes: 300.98 kWh of battery capacity, 1.89 kWp of PV power, and 0.79 kW of wind power.

In Table 4, the energy results show total PV production of 2863.84 kWh (70 %) and wind production of 1243.23 kWh (30%), thus meeting a demand of 3631.00 kWh with zero deficits. Furthermore, battery charging reaches 1452.25 kWh and discharging 1127.40 kWh, with total losses limited to 612.00 kWh (14.9% of production), mainly due to energy waste and battery losses. As a consequence, this indicates high efficiency, with renewable covering 100% of the demand.

Table 3 The results of sizing optimization

No. CBAT (kWh)

PPV (kWp)

PWT (kW)

Losses (kWh)

Execution Time (s)

1 300.98 1.89 0.79 612.00 949
2 300.98 1.89 0.79 612.00 964
3 301.20 1.89 0.78 612.09 944
4 300.98 1.89 0.79 612.00 933
5 300.98 1.89 0.79 612.00 941

Table 4 Results of energy balance of optimized system

Total Energy PV (kWh) Total Energy WT (kWh) Total Energy Demand (kWh) Total Energy Charged (kWh) Total Energy Discharged (kWh) Total Energy Loss (kWh) Total Energy Deficit (kWh)
2863.84 1243.35 3631.00 1452.25 1127.40 612.00 0

Figure 4 shows a stable convergence of the PSO algorithm, quickly reaching the minimum value of l.

image4.png

Fig. 4 The convergence of optimization

The relatively large battery capacity obtained by the optimization process (300.98kWh) is primarily related to the temporal mismatch between renewable energy generation and load demand over the annual operating horizon. The optimization objective aims to minimize system losses while ensuring reliable energy supply throughout the year. Consequently, the battery serves not only to compensate for short-term fluctuations in PV and wind generation but also to store surplus renewable energy during high-production periods for use during low-generation periods. Therefore, the resulting battery size reflects the need to accommodate seasonal and daily variations in renewable resources and load demand.

Figure 5 illustrates the evolution of the battery State of Charge (SOC) over an annual horizon. As shown in Figure 5, four phases can be observed in the seasonal behaviour of the battery SOC:

  • During the first months of the year (January–February), the SOC remains at a moderate level, varying between approximately 45% and 55%. This pattern indicates frequent charge–discharge cycles to ensure the energy balance between energy supply and load demand;

  • From early spring to mid-summer (approximately March to June), the SOC increases steadily and reaches values close to full capacity. This trend reflects a sustained energy surplus, mainly attributable to increased renewable energy production;

  • From mid-summer onward (July–September), the SOC starts to gradually decreas, indicating an increased reliance on the battery to compensate for reduced renewable generation or increasing energy demand;

  • By the end of the year (December), the SOC drops to relatively low levels, around 20–30%. This behaviour highlights long-term energy deficit conditions and a significant dependence on the battery storage system to ensure supply continuity.

image5.png

Fig. 5 Annual variation of the battery SOC under the hybrid energy system operation

image6.png

Fig. 6 Simulation of generated PV power profile

Figures 3 and 6 clearly show that peak demand periods coincide with peak solar irradiance levels. Furthermore, with a total production of 2863.84 kWh, the PV is considered the primary driver of the system, covering the majority of the demand.

image7.png

Fig. 7 Simulation of generated WT Profile

Figures 6 and 7 illustrate the simulated power profiles of the PV system and the WT, respectively, over the annual operating period. Moreover, the PV power profile exhibits pronounced intermittency with sharp peaks and frequent fluctuations, reflecting the strong dependence on solar irradiance variability due to diurnal cycles and changing weather conditions. In contrast, the WT power profile demonstrates a smoother and more sustained generation trend, with higher power levels maintained over extended periods, while still capturing the stochastic nature of wind speed variations. The complementary behaviour of PV and WT generation enhances the overall reliability of the hybrid system by reducing power deficits and improving the utilization of renewable resources throughout the year.

image8.png

Fig. 8 Power Balance Profiles for Four Representative Days (March, June, September, December)

Figure 8 shows hourly power balance profiles for four representative days in March, June, September, and December, illustrating the interaction between generated energy, load demand, and battery operation under different seasonal conditions. In spring and summer (March and June), higher renewable generation leads to surplus energy during midday hours, which is then stored in the battery, while the load is mainly supplied directly by the generation system. In autumn (September), a more balanced behaviour is observed, with moderate charging during peak generation periods and battery support during the evening hours. In winter (December), reduced generation results in limited battery charging and increased battery discharge to satisfy the load. Overall, these results highlight the effectiveness of the energy management strategy in maintaining power balance and ensuring supply continuity under different seasonal scenarios.

The optimized hybrid PV–wind–battery configuration exploits the complementary production patterns of solar and wind rather than merely adding capacities. In this regard, PV delivers most energy during daylight and is weather-dependent, while wind helps during mornings, evenings and seasons of low solar output. this complementarity reduces the temporal mismatch with the load, improves year-round renewable availability, and limits curtailment.

In terms of the battery, it serves both short-term smoothing and seasonal buffering: it stores surplus energy during high-generation periods and supplies energy during low-production seasons. Consequently, the seasonal SOC profile reflects charging in high-generation months and discharging when renewable input falls. This behaviour underlines the battery’s central role in energy balancing and supply continuity in an off-grid system.

However, battery losses must be weighed against the avoided curtailment; in this case, the trade-off is favourable: the battery recovers otherwise wasted renewable energy while keeping total losses limited, thus showing that the chosen capacity is large enough to absorb surpluses but not so large as to incur excessive idle losses or costs.

In practice, the optimized design ensures zero-deficit supply for the studied Skikda conditions, increases robustness to climate variability by reducing dependence on any single resource, and favours continuity and loss reduction over oversized generation. Therefore, this makes the loss-minimization sizing approach well suited to remote off-grid applications, and provides a physically meaningful guideline for real deployments, accordingly.

6. CONCLUSION

In light of the above findings, this work addresses the optimal sizing of a PV–Wind–ESS hybrid system using the PSO algorithm to determine the most suitable system configuration. The simulation results demonstrate that the proposed approach effectively achieves optimal design and sizing by identifying the best-performing solution for the hybrid energy system.

Moreover, the system delivers 100% reliable power (zero deficits) to meet an annual demand of 3631 kWh using minimal component sizes of 1.89 kWp PV, 0.79 kW WT and 300.98 kWh battery, with losses of just 14.9% of the total energy production of 4107.19 kWh.

Furthermore, the possibility of using this method has proven to be of great interest as it can be applied to different cases for the optimal sizing and management of hybrid renewable energy systems under different load profiles and meteorological conditions, as well.

In further work, other criteria will also be taken into consideration, such as Cost of Energy COE, Loss of Probability of Charge LPSP, with a focus on the study of the multi-objectives’ optimization to examine the relation between the total energy losses and other performance criteria.

Symbol

Description

Unit

PPV(t) Output power of a PV panel at time t W
PSTC Rated PV power at Standard Test Conditions (STC) W
G Incident solar irradiance on PV panel surface W/m²
GSTC Solar irradiance at Standard Test Conditions W/m²
γ Temperature power coefficient of the PV module °C⁻¹
Tc PV cell operating temperature °C
Tc,STC Cell temperature at Standard Test Conditions (25 °C) °C
Ta Ambient temperature °C
GNOCT Solar irradiance at Nominal Operating Cell Temperature (NOCT) conditions W/m²
Ta,NOCT Ambient temperature at NOCT conditions (20 °C) °C
NOCT Nominal Operating Cell Temperature (constant) °C
PPVnet(t) Net output power of PV generator at time t (after converter) W
Pw(t) Output power of wind turbine at time t W
Pn Rated power of wind turbine W
V Wind speed at hub height m/s
Vn Nominal (rated) wind speed (12.5 m/s) m/s
Vd Cut-in wind speed (2 m/s) m/s
Vc Cut-out wind speed (42 m/s) m/s
Pwnet(t) Net output power of wind turbine at time t (after converter) W
PB(t) Power exchanged with battery at time t (+ charging, − discharging) W
PBnet(t) Net power from/to battery (including losses) at time t W
PBloss(t) Battery efficiency-related power loss (charging or discharging) W
PBself-disc(t) Battery self-discharge power loss at time t W
WB(t) Stored energy in battery bank at time t Wh
QB Total capacity of battery bank Ah
VB Battery bank voltage V
SOC(t) State of Charge of battery at time t %
SOCMIN / SOCMAX Minimum / maximum allowable State of Charge (20% / 100%) %
PBmax,in(t) Maximum admissible charging power at time t W
PBmax,out(t) Maximum admissible discharging power at time t W
σ Self-discharge rate coefficient (0.004 % per hour) % h⁻¹
ηCONV DC/DC converter efficiency (0.96) —
ηB Battery round-trip efficiency (charging + discharging, 0.90) —
ηINV DC/AC inverter efficiency (0.96) —
PCONloss(t) Total power conversion loss in DC/DC converters at time t W
PTotalloss(t) Total power loss in the system at time t W
Pwasted(t) Surplus (wasted) power when battery is full and demand is met W
PLOAD(t) Load (demand) power at time t W
Tsim Simulation time step (1 hour) h
Xi(k) Position of particle i at iteration k —
Vi(k) Velocity of particle i at iteration k —
Xi,best Personal best position of particle i —
Xg,best Global best position found by the swarm —
c1, c2 Cognitive and social acceleration factors (both = 2.05) —
r1, r2 Uniformly distributed random numbers in [0, 1] —
w Inertia weight (dynamically updated, wmin = 0.4, wmax = 0.9) —
Niter Total number of PSO iterations (500) —
AES Autonomous Energy System —
COE Cost of Energy —
ESS Energy Storage System —
LCOE Levelized Cost of Energy —
LPSP Loss of Power Supply Probability —
NOCT Nominal Operating Cell Temperature —
PSO Particle Swarm Optimization —
PV Photovoltaic —
REF Renewable Energy Fraction —
SOC State of Charge —
STC Standard Test Conditions (G = 1000 W/m², Tc = 25 °C) —
WT Wind Turbine —

7. REFERENCES

1 

Sebbagh, T., Modelling, analysis, and techno-economic assessment of a clean power conversion system with green hydrogen production, Bull. Pol. Acad. Sci. Tech. Sci., Vol. 72, No. 4, p. e150115, 2024.https://doi.org/10.24425/bpasts.2024.150115

2 

Zegueur, A., Sebbagh, T., Metatla, A., Optimizing a sustainable power system with green hydrogen energy storage for telecommunication station loads, Solar Energy Sustain. Dev., Vol. 14, No. 1, pp. 1–19, 2025.https://doi.org/10.51646/jsesd.v14i1.366

3 

Kaabeche, A., Belhamel, M., Ibtiouen, R., Sizing optimization of grid-independent hybrid photovoltaic/wind power generation system, Energy, Vol. 36, No. 2, pp. 1214–1222, 2011.https://doi.org/10.1016/j.energy.2010.11.024

4 

Kapen, P.T., Multi-objective optimization of a Wind/Photovoltaic/Battery hybrid system using a novel hybrid meta-heuristic algorithm, Energy Convers. Manage., Vol. 327, 2025.https://doi.org/10.1016/j.enconman.2025.119533

5 

Huneke, F., Henkel, J., Benavides González, J.A., Erdmann, G., Optimisation of hybrid off-grid energy systems by linear programming, Energy Sustain. Soc., Vol. 2, No. 7, pp. 1–19, 2012.https://doi.org/10.1186/2192-0567-2-7

6 

Ekren, O., Ekren, B.Y., Size optimization of a PV/wind hybrid energy conversion system with battery storage using response surface methodology, Appl. Energy, Vol. 85, No. 11, pp. 1086–1101, 2008.https://doi.org/10.1016/j.apenergy.2008.02.016

7 

Ekren, O., Ekren, B.Y., Ozerdem, B., break-even analysis and size optimization of a PV/wind hybrid energy conversion system with battery storage – A case study, Appl. Energy, Vol. 86, No. 7–8, pp. 1043–1054, 2009.

https://doi.org/10.1016/j.apenergy.2008.09.024

8 

Nguyen, T.A., Crow, M.L., Optimization in energy and power management for renewable-diesel microgrids using dynamic programming algorithm, Proc. IEEE Int. Conf. Cyber Technol. Autom. Control Intell. Syst. (CYBER 2012), Bangkok, Thailand, May 27–31, 2012, pp. 11–16.https://doi.org/10.1109/CYBER.2012.6392518

9 

Sebbagh, T., Kelaiaia, R., Abdelouahed, A., Zaatri, A., Optimizing the use of green energies, an application to crop irrigation, Carpathian J. Electr. Eng., Vol. 12, No. 1, 2018.

10 

Rouhani, A., Jabbari, M., Honarmand, S., A teaching learning based optimization for optimal design of a hybrid energy system, Int. J. Electr. Comput. Energy Electron. Commun. Eng., Vol. 9, No. 8, pp. 737–744, 2015.

11 

Yang, H., Zhou, W., Lu, L., Fang, Z., Optimal sizing method for stand-alone hybrid solar–wind system with LPSP technology by using genetic algorithm, Solar Energy, Vol. 82, No. 4, pp. 354–367, 2008.https://doi.org/10.1016/j.solener.2007.08.005

12 

Amer, M., Namaane, A., M’Sirdi, N.K., Optimization of hybrid renewable energy systems (HRES) using PSO for cost reduction, Energy Procedia, Vol. 42, pp. 318–327, 2013.https://doi.org/10.1016/j.egypro.2013.11.032

13 

Katsigiannis, Y.A., Georgilakis, P.S., Optimal sizing of small isolated hybrid power systems using tabu search, J. Optoelectron. Adv. Mater., Vol. 10, No. 5, pp. 1241–1245, 2008.

14 

Ekren, O., Ekren, B.Y., Size optimization of a PV/wind hybrid energy conversion system with battery storage using simulated annealing, Appl. Energy, Vol. 87, No. 2, pp. 592–598, 2010.https://doi.org/10.1016/j.apenergy.2009.05.022

15 

Katsigiannis, Y.A., Georgilakis, P.S., Karapidakis, E.S., Hybrid simulated annealing–tabu search method for optimal sizing of autonomous power systems with renewables, IEEE Trans. Sustain. Energy, Vol. 3, No. 3, pp. 330–338, 2012.

https://doi.org/10.1109/TSTE.2012.2184840

16 

Katsigiannis, Y.A., Kanellos, F.D., Papaefthimiou, S., A software tool for capacity optimization of hybrid power systems including renewable energy technologies based on a hybrid genetic algorithm tabu search optimization methodology, Energy Syst., Vol. 7, No. 1, pp. 33–48, 2014.https://doi.org/10.1007/s12667-014-0138-0

17 

Gabour, A., Metatla, A., Kelaiaia, R., Bourennani, F., Karboua, A., Recent bibliography on the optimization of multi-source energy systems, Arch. Comput. Methods Eng., Vol. 26, No. 4, pp. 809–830, 2019.https://doi.org/10.1007/s11831-019-09371-7

18 

Bouakkaz, A., Gil Mena, A.J., Haddad, S., Ferrari, M.L., Efficient energy scheduling considering cost reduction and energy saving in hybrid energy system with energy storage, J. Energy Storage, Vol. 33, Art. 101887, 2021.

https://doi.org/10.1016/j.est.2020.101887

19 

Kennedy, J., Eberhart, R., Particle Swarm Optimization, Proc. IEEE Int. Conf. Neural Networks (ICNN’95), Perth, Australia, Vol. 4, pp. 1942–1948, 1995.


This display is generated from NISO JATS XML with jats-html.xsl. The XSLT engine is libxslt.